Meta Platforms announced on Tuesday, November 17, 2026, that it has upgraded Instagram’s content recommendation algorithms in India using more advanced artificial intelligence (AI) models to sharply restrict the reach of sexually suggestive material and violent content.
The updated AI filters specifically target posts suggested from accounts that users do not follow across Reels, Explore, and Feed. Meta clarified that the measures govern algorithmic distribution rather than imposing an outright platform-wide ban on content that does not violate its core Community Standards—and noted that user reports on recommended content in India have dropped by half over the past year.
Regulatory Scrutiny & How the New AI Safeguards Work
The rollout comes a month after India’s Ministry of Electronics and Information Technology (MeitY) issued an October 10 advisory directing Meta to review its ranking and recommendation algorithms under the IT Rules, 2021, so that manipulated, misleading, or out-of-context posts capable of inciting violence or disrupting public order are not disproportionately amplified:
- Proactive Detection Against Recommendations Guidelines: Unlike Instagram’s Community Standards—under which prohibited or unlawful content is removed outright—its Recommendations Guidelines apply a stricter threshold for surfacing posts to wider audiences. The upgraded AI models evaluate posts proactively before users even have to hide or report them.
- Enhanced User Controls & Teen Protections: Alongside automated AI filters, Meta highlighted user-level tools available in India, including Sensitive Content Control, options to limit recommended political posts, Hidden Words filters for comments and message requests, and Teen Accounts—which automatically place teenage users into stricter 13-plus age-appropriate content settings.













